get_ipython().magic("time print(np.sqrt(-cross_val_score(model, trainingData, trainingData['isSpam'], cv=10, scoring='mean_squared_error')).mean())")
return
bestRun = []
for _ in range(20):
rand = RandomizedSearchCV(model, paramDistribution, cv=10, scoring = 'accuracy', n_iter = 10)
rand.fit(trainingData, trainingData['isSpam'])
# examine the best model
bestRun.append({'score' : round(rand.best_score_,3), 'params' : rand.best_params_})
print(max(bestRun, key=lambda x:x['score']))
return max(bestRun, key=lambda x:x['score'])
#read site
rawSite = read_mongo(db = 'CB', collection = 'site', host = 'localhost', no_id = False)
siteModified = rawSite.drop(['dismissedOnboarding', 'feedCounter', 'feedToken',
'modules', 'password', 'theme', 'photoId', 'requestAccess',
'requestPassword', 'bi', 'photo', 'goFundMe', 'lastName',
'numAmps', 'partner', 'size', 'theme', 'createFormSessionId', 'allowList',
'blockList', 'displayEmail', 'isPhotoOrderingFixed', 'healthCondition',
'spam', 'status', 'firstName', 'lastInvite', 'isDeleted',
'hasCommentFix','age'], axis = 1)
viewPort = siteModified.cm.apply(pd.Series).fillna(-1)
siteModified['hasJavaScriptOn'] = [0 if vp == -1 else 1 for vp in viewPort.vpw]
siteModified.drop(['cm'], axis = 1, inplace = True)
siteModified['descriptionLen'] = rawSite.description.str.len()
siteModified.drop(['description'], axis = 1, inplace = True)
siteModified['nameLen'] = rawSite.name.str.len()
siteModified.drop(['name'], axis = 1, inplace = True)
siteModified['titleLen'] = rawSite.title.str.len()
final = pd.concat([test, dfWithClass], axis=1)
#take a look at the confusion matrix
print(pd.crosstab(final.isSpam, final.predictedClass))
print("0s: %d, 1s: %d" %
(np.sum((final.isSpam == 0) & (final.predictedClass == 0)),
np.sum((final.isSpam == 1) & (final.predictedClass == 1))))
print(
"Accuracy: %.3f" %
float(np.sum(final.isSpam == final.predictedClass) / float(len(test))))
print("Precision: %.3f" % float(
np.sum((final.isSpam == 1) & (final.predictedClass == 1)) /
np.sum(final.isSpam == 1)))
#read journals
rawJournals = read_mongo(db='CB', collection='journal', host='localhost')
journals = pd.DataFrame(list(rawJournals['body']), columns=['content'])
journals['siteId'] = rawJournals['siteId']
journals['text'] = rawJournals['title'].astype(str) + ' ' + journals['content']
journals.drop(['content'], inplace=True, axis=1)
#read siteIds
rawSite = read_mongo(db='CB', collection='site', host='localhost', no_id=False)
siteIds = pd.DataFrame(list(rawSite['_id']), columns=['siteId'])
siteIds['isSpam'] = rawSite['isSpam']
siteIds.isSpam.fillna(0, inplace=True)
siteIds.rename(columns={'isSpam': 'isSiteSpam'}, inplace=True)
#spam data from file
octSiteProfileSpam = pd.read_csv(
"/Users/dmurali/Documents/spamlist_round25_from_20150809_to_20151015.csv",
paramDistribution,
cv=10,
scoring='accuracy',
n_iter=10)
rand.fit(trainingData, trainingData['isSpam'])
# examine the best model
bestRun.append({
'score': round(rand.best_score_, 3),
'params': rand.best_params_
})
print(max(bestRun, key=lambda x: x['score']))
return max(bestRun, key=lambda x: x['score'])
#read site
rawSite = read_mongo(db='CB', collection='site', host='localhost', no_id=False)
siteModified = rawSite.drop([
'dismissedOnboarding', 'feedCounter', 'feedToken', 'modules', 'password',
'theme', 'photoId', 'requestAccess', 'requestPassword', 'bi', 'photo',
'goFundMe', 'lastName', 'numAmps', 'partner', 'size', 'theme',
'createFormSessionId', 'allowList', 'blockList', 'displayEmail',
'isPhotoOrderingFixed', 'healthCondition', 'spam', 'status', 'firstName',
'lastInvite', 'isDeleted', 'hasCommentFix', 'age'
],
axis=1)
viewPort = siteModified.cm.apply(pd.Series).fillna(-1)
siteModified['hasJavaScriptOn'] = [0 if vp == -1 else 1 for vp in viewPort.vpw]
siteModified.drop(['cm'], axis=1, inplace=True)
siteModified['descriptionLen'] = rawSite.description.str.len()
siteModified.drop(['description'], axis=1, inplace=True)
siteModified['nameLen'] = rawSite.name.str.len()
predictor.fit(train, train['isSpam'])
predicted = predictor.predict(test)
dfWithClass = pd.DataFrame(predicted, columns = ['predictedClass'])
final = pd.concat([test, dfWithClass], axis=1)
#take a look at the confusion matrix
print(pd.crosstab(final.isSpam, final.predictedClass))
print("0s: %d, 1s: %d" %(np.sum((final.isSpam == 0) & (final.predictedClass == 0)), np.sum((final.isSpam == 1) & (final.predictedClass == 1))))
print("Accuracy: %.3f" %float(np.sum(final.isSpam == final.predictedClass) / float(len(test))))
print("Precision: %.3f" %float(np.sum((final.isSpam == 1) & (final.predictedClass == 1)) / np.sum(final.isSpam == 1)))
#read journals
rawJournals = read_mongo(db = 'CB', collection = 'journal', host = 'localhost')
journals = pd.DataFrame(list(rawJournals['body']), columns = ['content'])
journals['siteId'] = rawJournals['siteId']
journals['text'] = rawJournals['title'].astype(str) + ' ' + journals['content']
journals.drop(['content'], inplace = True, axis = 1)
#read siteIds
rawSite = read_mongo(db = 'CB', collection = 'site', host = 'localhost', no_id = False)
siteIds = pd.DataFrame(list(rawSite['_id']), columns = ['siteId'])
siteIds['isSpam'] = rawSite['isSpam']
siteIds.isSpam.fillna(0, inplace = True)
siteIds.rename(columns = {'isSpam':'isSiteSpam'}, inplace = True)
#spam data from file
octSiteProfileSpam = pd.read_csv("/Users/dmurali/Documents/spamlist_round25_from_20150809_to_20151015.csv",
usecols = ['siteId','isSpam'])
